Artificial Intelligence & Machine Learning , Governance & Risk Management , Next-Generation Technologies & Secure Development

Meta Tests Facial Recognition to Curb Deepfake Scams

Firm Won't Deploy Feature in the EU, UK Due to Data Collection Norms
Meta Tests Facial Recognition to Curb Deepfake Scams
Image: Shutterstock

Meta is rolling out facial recognition technology on its social media platforms to spot scam ads featuring celebrity deepfakes.

See Also: Establishing a Governance Framework for AI-Powered Applications

The feature also aims to verify the identities of users locked out of their Facebook or Instagram accounts by having them take selfie videos and comparing them with their profile pictures, said Monika Bickert, Meta's vice president of content policy. Users can upload photos of their government-issued ID cards for the same purpose.

Called "celebrity bait ads," deepfake scams use AI-generated images of popular personalities to trick users into parting with their money. Meta plans to use the facial recognition tool to compare the pictures in the scam ads with the celebrities' profile pictures to confirm their veracity. Only a "small group of celebrities" are enrolled in the program now, Bickert said.

It announced changes to how it labels content it suspects of being generated by AI in July, displaying an "AI Info" label for such content.

Meta earlier this month said it took down 8,000 of the "celeb bait" scam ads from the social media platforms after receiving 102 reports since April from the Australian Financial Crimes Exchange.

The company will not deploy the feature in the European Union and the United Kingdom because they would violate local data collection laws, Bickert said.

This is the company's first major expansion of its facial recognition system in a decade, after government investigations, privacy blowback and regulatory concerns forced it to roll back a "tagging" feature on Facebook that automatically identified people in photos. It deleted the facial data of more than a billion users. The company said at the time that it still saw facial recognition technology as a "powerful tool" to verify people's identity and prevent fraud and impersonation.

Bickert said the company partnered with policymakers, regulators and privacy and security experts before starting the latest tests. Despite making headlines several times for its lax data collection and storage practices, the company did not detail how it would manage security and privacy in its latest facial recognition project.

The company plans to roll out the facial recognition feature for account recovery to all users in the next few months, depending on the results of the beta tests. It will expand the celebrity bait ad tests in December to another group of celebrities.


About the Author

Rashmi Ramesh

Rashmi Ramesh

Assistant Editor, Global News Desk, ISMG

Ramesh has seven years of experience writing and editing stories on finance, enterprise and consumer technology, and diversity and inclusion. She has previously worked at formerly News Corp-owned TechCircle, business daily The Economic Times and The New Indian Express.




Around the Network

Our website uses cookies. Cookies enable us to provide the best experience possible and help us understand how visitors use our website. By browsing databreachtoday.com, you agree to our use of cookies.